Recent advances in the technology of structured psychiatric evaluation have proved that reliable psychiatric diagnoses can be made in preschool age (2-5 years) children using the DSM-IV diagnostic criteria. In a relatively small earlier study of psychopathology in general, we found that the prevalence of anxiety disorders in unselected pediatric clinic attendees was 9%. We now propose to conduct the first specific study of DSM-IV anxiety disorders in a preschool pediatric population. We plan to examine the validity of such disorders in terms of their relationships with (1) disability (psychosocial impairment), (2) associations with previously validated scale measures of psychopathology, (3) teacher/caregiver reports of problems, (4) temperamental extremes, (5) measures of attention and inhibitory control, and (6) parental history of psychopathology. We will also determine whether the DSM-IV nosology provides an adequate description of the patterns of association among anxiety symptoms in this age group. We plan to screen an initial sample of 4,900 unselected attendees at several Duke University Medical Center affiliated non-specialty pediatric clinics, using a parent-report anxiety disorders screening measure we have developed. We will then conduct parent-reported interviews about the child for those scoring in the top 20% on this screen, plus a 7.9% random sample of the rest (for a total of 929 completed interviews). From these interviews we will identify nested groups of 250 anxious disorder cases and 250 non-anxious controls for more intensive study, including teacher/caregiver assessments and intensive laboratory assessment with the parent and child. If, as we expect, these diagnoses turn out to be valid, the study will have enormous implications for pediatric clinical practice, because we know that they are not currently being identified, referred for specialist evaluation, or treated by pediatricians. It will also indicate that "prevention" efforts for anxiety disorders will need to be started much earlier than is usually supposed.